Zuckerberg is pouring billions into plans for 'personal superintelligence'

Zuckerberg is pouring billions into plans for ‘personal superintelligence’

Meta (META) CEO Mark Zuckerberg is plowing billions into his company’s AI expansion efforts with the goal of developing so-called superintelligence, or AI that can surpass human capabilities.

In a Threads post on Monday, Zuckerberg revealed plans to spend hundreds of billions of dollars building several massive AI data centers across the US, including one that will come online as soon as next year.

Another data center, called Hyperion, will eventually scale up to support up to 5 gigawatts, or 5 billion watts, of capacity. A gigawatt of electricity can power roughly 800,000 homes.

It’s not just data centers, though. Zuckerberg is also splashing money on high-profile AI hires. In June, Meta invested $14.3 billion in Scale AI and hired its CEO Alexandr Wang. The company also hired former GitHub CEO Nat Friedman and Safe Superintelligence CEO Daniel Gross and poached Apple’s head of AI foundation models, Ruoming Pang, according to Bloomberg.

According to the Wall Street Journal, Meta also snagged OpenAI (OPAI.PVT) researchers Lucas Beyer, Alexander Kolesnikov, and Xiaohua Zhai.

All of this comes as Meta looks to capture the lead in the AI wars after the company was forced to put its Llama 4 Behemoth model on ice while it works to bring its functionality up to par with frontier models from the likes of OpenAI and Google.

“Mr. Zuckerberg believes it’s very important to have one of the leading models,” D.A. Davidson head of technology research Gil Luria told Yahoo Finance.

“He’s rebuilt the team so he has the talent to do it, and he’s going to continue to invest in the compute capacity so he has the compute capacity to achieve it.”

For Meta, it’s a gamble the company can’t afford to lose.

Meta’s AI efforts differ significantly from those at OpenAI, Anthropic (ANTH.PVT), Perplexity (PEAI.PVT), xAI, and even Microsoft (MSFT) and Google (GOOG, GOOGL). Rather than using its technology to explore questions about the universe or power enterprise and cloud-based software services, Zuckerberg said the company is angling to make its AI platform a kind of “personal superintelligence.”

During an interview with The Information’s TITV, Zuckerberg explained that Meta’s personal superintelligence will solve simpler problems for users on a daily basis.

“Our mission with the lab is to deliver personal superintelligence to everyone in the world, so that way we can put that power in every individual’s hand. It’s a different thing than what the other labs are doing. This is going to be something that is the most important technology in our lives.”

It’s a vague statement, with no real solid examples of how personal superintelligence will impact our daily lives or how we’ll use it more broadly. But Zuckerberg did offer that the technology will power Meta’s recommendation engine, advertising capabilities for customers, and, importantly, the company’s smart glasses software.

“I continue to think that that’s going to be the best form factor for AI, because, you know, they can see what you see and hear what you hear, and you can talk to them throughout the day,” Zuckerberg said.

Meta is banking on smart glasses as the next major piece of consumer hardware after the smartphone. The company is developing self-contained smart glasses that don’t require a smartphone to power apps or other software. The move would allow Meta to break free of Apple’s and Google’s respective app stores, giving it more control over its products, a long-term goal of the social media giant.

In order for Meta to get to that point, though, it needs to spend a lot of cash — something it’s doing with aplomb.

“Overall, I think it just shows how in demand AI engineering talent is and how hard it is to secure and keep good talent,” Creative Strategies principal analyst Ben Jajarin told Yahoo Finance.

“The AI talent wars is a fierce battle, and honestly, it is hard to see this going away or slowing down anytime soon,” he said. “Ultimately, other than money, I think it will be hard to keep talent as well. So it’s possible we will see a lot of churn as well.”

The spending on talent comes after as Meta looks to develop future cutting-edge AI models following its Llama 4 issues.

“Last year, when they introduced the Llama 3 model, they had one of the top frontier models at the time,” Luria said.

“However, by the time they got around to Llama 4 it was not successful, and it’s not even one of the best models,” he added. “And so Mr. Zuckerberg decided that it was essential for him to redouble his effort and remake the team leading the AI effort at Meta, which he has done in an unprecedentedly aggressive manner over the last few weeks.”

Now Zuckerberg has to ensure all of that spending pays off and that his vision of a more personal superintelligence comes to fruition.